The system hit another snag at 244 MHz, this time we had to give the memory a big voltage increase, all the way up to 3.2V. That did not just fix the problem it also allowed the memory to crack the 250 MHz barrier, with 2-2-2-5 memory timings, which is not bad at all! With 3.2V the memory went as high as 257 MHz before I started to experience stability issues again; the fix was simple though, I just raised the voltage further.